120Hz hum, new caps, old tube unit, chassis-ground questions..

[Speedskater]

The Earth Ground is the AC power Safety Ground/Protective Earth and is correctly connected to the chassis near where the cord enters.

The chassis is acting as a Ground Plane for the audio circuit Common and DC supply common.

I would suspect one or more of those connecting ground screws is making a poor connection to the chassis.

 

[ruffrecords]

OK, I just wanted to be sure the chassis was connected properly to safety earth.

The grounding scheme is not brilliant but I don't think it is like to be the issue. The fact that you hear 120Hz hum leads me to suspect the HT smoothing is the issue. The main reservoir capacitor C12B is only 40uF so there will be considerable ripple at that point. For example it the HT current draw is 50mA then the ripple will be 10V peak to peak.

Vripple = Idc/(f * C)

In the original design, three things act to reduce this. First the push pull output stage tends to cancel ripple. Second the 800 ohm plus 40uF (C12C) RC filter for the screens and the 6SN7 will drop the ripple there by about 28dB. And third, the negative feedback will reduce the ripple in the output by the amount of feedback which may well contribute another 20dB of ripple reduction.

The one thing you have changed is the negative feedback which you have reduced so as to increase the gain so that modification will certainly have worsened the amount of ripple in the output. The other factor is the PSRR of the 6SN7 stage is not brilliant so some ripple will undoubtedly creep in there..

There is nothing you can do about the main reservoir capacitor C12B because it is determined by the 5AR4 rectifier tube. But you could increase C12C significantly to say 220uF and see if that measurably reduces the 120Hz hum.

Cheers

Ian

 

[systemtruck]

OK, I just wanted to be sure the chassis was connected properly to safety earth.

The grounding scheme is not brilliant but I don't think it is like to be the issue. The fact that you hear 120Hz hum leads me to suspect the HT smoothing is the issue. The main reservoir capacitor C12B is only 40uF so there will be considerable ripple at that point. For example it the HT current draw is 50mA then the ripple will be 10V peak to peak.

Vripple = Idc/(f * C)

In the original design, three things act to reduce this. First the push pull output stage tends to cancel ripple. Second the 800 ohm plus 40uF (C12C) RC filter for the screens and the 6SN7 will drop the ripple there by about 28dB. And third, the negative feedback will reduce the ripple in the output by the amount of feedback which may well contribute another 20dB of ripple reduction.

The one thing you have changed is the negative feedback which you have reduced so as to increase the gain so that modification will certainly have worsened the amount of ripple in the output. The other factor is the PSRR of the 6SN7 stage is not brilliant so some ripple will undoubtedly creep in there..

There is nothing you can do about the main reservoir capacitor C12B because it is determined by the 5AR4 rectifier tube. But you could increase C12C significantly to say 220uF and see if that measurably reduces the 120Hz hum.

Cheers

Ian

Ok great. This all makes sense, thank you!
I have a handful of 100uF 450v caps arriving soon and will tack a couple on to C12C in parallel. I’m sure it will help because even just adding 80uF already sounds way better.

The tube installed is a 5Y3. The screen printing on the chassis at the tube location says both.. 5AR4 / 5Y3. I checked the datasheet for 5Y3 and was surprised to see 10uF listed as max capacitance of first cap. So it’s at 4X the value from factory.

Do you suppose a 5AR4 would improve ripple? I know it will change some things. I don’t see a max cap value in the 5AR4 datasheet that’s on Franks, but an Amperex 5AR4 datasheet has a max 60uF listed. I DO like the sag sound that the 5Y3 has though.

And on that note… would increasing the C12C to 220uF change any performance like sag or frequency response?

I may just pop in a 5AR4 and see if it improves 120Hz hum. It’s more important to me to have hum go down than have the perfect amp sag sound.

Any thoughts about C12A? I’m going to use that as the filter for my preamp. But it is two tubes, so four stages, so maybe i should create yet another filter stage so that the first half of preamp has that one and the second half of preamp has the C12A stage. I would also like to lower the voltage for the total preamp circuit, so maybe I’ll increase R28 some as well. That should i suppose increase the filtering some more too which would be nice.

 

[ruffrecords]

Changing to a 5AR4 may not make much difference. I think it has a higher current rating which means a lower series resistance which could actually make the ripple worse.

increasing C12C will have minimal effect on frequency response and will not affect sag; only C12B does that.

You can certainly increase C12A.

I cannot find R28.

Cheers

Ian

 

[Tubetec]

Heres the later Hp200cd for comparrison , they added a choke ,

I got a few good pointers from former member PRR about the 200cd , a search should reveal the topic .

One thing he impressed on me was Hewlett and Packard didnt cut corners and didnt leave things to chance , these guys were designing the cutting edge tech of the time ,

Dont try and second guess the layout and grounding until you've checked everything
It was alluded to already ,but its good to slacken off any chassis through bolts that form ground with the various parts of the circuit , inspect and as nessesary remove any rust or oxidation from the nuts, bolts, washers, solder tags and chassis , then re-torque everything back down to form a new gas tight metal on metal contact .

I ended up using a foil motor run cap of 8uf in place of 10uf as C12B , that allowed me hit the target HT volts accurately . You can easily lower the voltage at the rectifier output by selecting a lower value than 40uf as first filter .

 

[systemtruck]

Heres the later Hp200cd for comparrison , they added a choke ,

I got a few good pointers from former member PRR about the 200cd , a search should reveal the topic .

One thing he impressed on me was Hewlett and Packard didnt cut corners and didnt leave things to chance , these guys were designing the cutting edge tech of the time ,

Dont try and second guess the layout and grounding until you've checked everything
It was alluded to already ,but its good to slacken off any chassis through bolts that form ground with the various parts of the circuit , inspect and as nessesary remove any rust or oxidation from the nuts, bolts, washers, solder tags and chassis , then re-torque everything back down to form a new gas tight metal on metal contact .

I ended up using a foil motor run cap of 8uf in place of 10uf as C12B , that allowed me hit the target HT volts accurately . You can easily lower the voltage at the rectifier output by selecting a lower value than 40uf as first filter .




View attachment 125050

Changing to a 5AR4 may not make much difference. I think it has a higher current rating which means a lower series resistance which could actually make the ripple worse.

increasing C12C will have minimal effect on frequency response and will not affect sag; only C12B does that.

You can certainly increase C12A.

I cannot find R28.

Cheers

Ian

Amazing. It’s really inspiring to hear this vote of confidence for HP.
I’ll take this advice and make sure hard metal connections are as good as they can be.
Once I get this 120Hz sounding as quiet as it can in the context of not having negative feedback, I will be gutting the oscillator section and installing the preamp, which hopefully can be pretty quiet.

As for lowering the first cap to decrease output voltage, wow I didn’t know that!

R28 is up with the C12A in schematic, as part of that RC filter.

The distortion sounds from this output without NFB is just otherworldly. I’ve never heard anything sound like this. I will put in a recently tested pair of RCA 6V6’s and adjust the power tube distortion/offset pot to see if I can bring down the hum. But I will be marking the offset pot where it currently is. Although I doubt the sound of the overdrive will change from this pot, since it’s not quite the “distortion” HP was intending to adjust here haha.

 

[Tubetec]

I have a couple of smaller amps using the split load phase invertor , theres definately an interesting growl to the sound when overdriven ,

As I said previously I have an amp I made where the boost mode cuts the nfb from transformer secondary back to phase invertor , it not only boosts but changes the character of the amps distortion ,

 

[ruffrecords]

Looks like HP changed quite a few other things in that later design. The tube rectifier is gone and replaced by semiconductors, the output tubes are different and the NFB scheme appears to have been expanded to allow variable gain.

Cheers

Ian

 

[Tubetec]

The HP200 cd had a few revisions over its long life ,
I have the version seen above with 6AU6 and EL86 ,but with a GZ34 rectifier .

I dont think it varies FB , the circuit runs at full amplitude all the time , it has an attenuator across the output . It also switches to a different transformer for the 500khz range .

 

[systemtruck]

https://www.hpmemoryproject.org/wb_pages/wall_b_page_10a.htm

The one major thing lacking on the 201C is the balanced line output, but as far as i can tell it would be pretty easy to bypass the attenuator and utilize both top and bottom of the transformer secondary and that’s that, and also add a pair of 1200 ohm shunt resistors to ground on each leg of secondary?

 

[Walter66]

The input pentode 6AQ5 has 310V on the screen grid without any resistor to screen grid. Bring that voltage down to 80V with a resistor. The cathode resistor of that tube has 1500pF cap parallel to resistor, which is unlikely for an audio amp. Put in 50uF. Trim the voltage gain of that stage with the added screen grid resistor. Maybe that's why there is hum on the amp.

 

[systemtruck]

The input pentode 6AQ5 has 310V on the screen grid without any resistor to screen grid. Bring that voltage down to 80V with a resistor. The cathode resistor of that tube has 1500pF cap parallel to resistor, which is unlikely for an audio amp. Put in 50uF. Trim the voltage gain of that stage with the added screen grid resistor. Maybe that's why there is hum on the amp.

Thanks Walter! I’m actually deleting that whole 6AQ5/6AU6 section which is the oscillator. It’s currently disabled by removing the light bulb, as instructed in service manual, because there is some bleed in the background. However, I wonder if maybe when the osc tubes are active, which they are, that maybe they contribute somehow to 120Hz hum. It’s worth noting that the 120Hz hum has nothing to do with the oscillation itself, since it remains no matter what frequency the osc is set to.

 

[Walter66]

Thanks Walter! I’m actually deleting that whole 6AQ5/6AU6 section which is the oscillator. It’s currently disabled by removing the light bulb, as instructed in service manual, because there is some bleed in the background. However, I wonder if maybe when the osc tubes are active, which they are, that maybe they contribute somehow to 120Hz hum. It’s worth noting that the 120Hz hum has nothing to do with the oscillation itself, since it remains no matter what frequency the osc is set to.

Yes, set oscillator out of function. Then mod the amp to practical values of gain for an audio amp, measure where hum is induced (PSU or amp section). If PSU, change components/values/ improve circuit (added choke etc.). If amp section fails, find the topology - or circuit error. Good luck!

 

[Tubetec]

6AQ5 is a power output pentode , its not much use for gain but it will swing lots of volts when fed by the 6AU6

If you want to run the O/P stage without feed back , you wont need all that gain or voltage drive capabillity , but it would allow you incorporate a passive tone stack if you want .

Duncans PSUD software is great for doing a virtual mock up of your HT supply and allows you substitute in other values and anticipate what voltages to expect with good accuracy .

Duncans TSC might also be of interest .

It might be a help if you reposted a coherent version of the schematic because theres elements missing .

 

[systemtruck]

Sorry about that. Here’s the whole deal below. Before i had used screen shots and didn’t notice the top was clipped because it’s the oscillator and i haven’t been paying attention to that area.
That software sounds promising. Thank you! I will check that out for sure.
When I delete the oscillator section, i will have a preamp that has room for a passive tone stack i think. Inevitably, given that this is a guitar amp, I’ll want to try out some tone stacks!

 

[systemtruck]

Interesting development..
The R48 pot, between cathodes of power tubes, doesn’t exist. Naturally R47 from wiper to ground doesn’t exist either.
The pins 8 of each 6v6 are wired together.
So I checked the service manual list of changes and sure enough found a change, from the above, to a pair of parallel resistors to ground.
And, those resistors are in my unit coming from one of the tubes to ground.
So I suppose HP knows best? Maybe they installed only perfectly matching 6v6’s and thought to hell with this adjustment?
I installed my solid testing pair of 6v6’s I bought and the hum is identical. Put the HP 6v6’s back in.

So I guess I could either install a R48, R47, or I could just rule out this adjustment from the list of potential hum busters.
I’m bummed, was looking forward to seeing if it helped!

 

[Tubetec]

A cathode bias balance pot does help reduce hum from my experience ,

This schematic shows another variation ,

 

[systemtruck]

Below is a recording of 4 cap values at C12C

Transformer secondary bottom leg is 4700 to ground
NFB path is killed
V3 pin 6 has 4700 to ground
Oscillator audio out is cut, input to C9 is grounded.
Osc bulb is removed to kill oscillation, but tubes and circuits are still powered / active

https://www.dropbox.com/scl/fi/12qs....340.aif?rlkey=m8kszbnkd0zv2kbxisdl2mqq7&dl=0

There is white noise, which is the noise floor of the audio interface itself. I had turned down the attenuator of the 201C output and forgot to turn it back up when recording hum. Just ignore the white noise.

40uf, 120Hz is dominant, mains hum is underneath
140uf, 120Hz is almost inaudible and mains hum is now clear
240uf, 120Hz is gone, mains hum the same
340uf, no audible change, mains hum the same

I now think that the DC rail hum is eliminated enough, with 240uF. I’m convinced that transformer proximity is the remaining mains hum. Because I have control over the 120Hz now, and this doesn’t change the mains hum. If this is the case, I don’t think I can install a preamp inside the unit. I think it will pick up a lot of mains hum. HP designed this chassis with an extremely loud oscillator inside, with an output that I measured as 17 VAC. If it doesn’t pick up much mains hum, it is certainly loud enough to overpower it in the output stage. A preamp could over power the output mains hum too but I think it’ll pick up more mains hum than an oscillator would.

I suppose the only solutions are to relocate the power transformer to outside the unit in its own enclosure, or to install some steel barriers around power transformer.

I have a perfect working regulated tube variable power supply. I am considering disconnecting the 201C’s power transformer and patching in the power from the external supply to see if the mains hum is more or less eliminated. If it is, then I have all the info I need to finish this project! I know an external power chassis adds complications but it might be worth it.

 

[Tubetec]

I did a single tube pre-amp in a small aluminium box with a short wired in umbilical cable to the power amp enclosure for heaters and HT ,
Signal is on jacks in the usual way ,
With around 25 cm of cable and careful placement I could avoid 60hz field from the mains transformer entirely in the first pre amp stage ,
The other thing you can avoid is vibrations from the mains transformer being induced microphonically through the chassis into the tube ,
That becomes an issue at high gain ,
Let alone speaker cab vibrations ,

 

[Tubetec]

Im just looking back at things now again ,
You could make the 6AQ5 stage a reverb tank driver with the usual transformer .